Sistemas hidrológicos em transformação: impactos das mudanças antrópicas na dinâmica hidrológica da bacia hidrográfica do Rio São Francisco, Brasil

Abstract

In the Anthropocene, the increasing human influence on the hydrological cycle of watersheds has made their water availability and response patterns to climate extremes – both wet and dry – depend on both natural and anthropogenic factors. In this context, this thesis evaluated the impact of anthropogenic changes on the hydrological dynamics of watersheds. As a case study, the São Francisco River Watershed (BHSF) and the Grande River Basin (GRB), an important sub-basin of the BHSF, were used. Three different approaches were employed for this proposed assessment: the first applied more generally to the BHSF, while the second and third were applied more specifically to the GRB. In the first approach, the concept of climate elasticity of streamflow was applied to 10 sub-basins of the BHSF, enabling an evaluation of streamflow sensitivity to climate variations and the decomposition of total streamflow variation into climatic and anthropogenic components. In the second, a new methodology for assessing the impact of human activities on the development of hydrological droughts was proposed: the Drought Roulette Analysis (DRA). In the third, the watershed was interpreted as a complex hydrological system with multiple equilibrium states that switch between them in response to climatic and anthropogenic forcing. The results from these approaches led to the conclusion that anthropogenic activities have decreased water availability, increased the sensitivity of streamflow to climate variations, and slowed hydrological recovery after drought events, with streamflow deviations persisting even after precipitation levels are restored. Equally important,

the results from the third approach showed that a watershed can naturally exhibit a non- stationary hydrological regime characterized by multiple hydrological states and a variable

rainfall-runoff relationship; anthropogenic activities, in turn, modulate this non-stationarity, accentuating the asymmetry in transitions between these states. Additionally, incorporating information on hydrological states significantly improved the hydrological modeling process, providing a more dynamic representation of the variable rainfall-runoff relationship. Overall, these findings point to a dynamic risk to water availability and hydrological extremes in the BHSF, and more specifically in the Grande River watershed. This dynamic risk arises from the non-stationarity of the hydrological regime, which, although it can be caused by natural factors (climate variability), is strongly linked to human interventions induced in these watersheds. These interventions, in addition to potentially generating non-stationarity, modify transitions between hydrological states and the patterns of drought evolution. The aforementioned dynamic risk introduces significant uncertainty into the water resource management and planning process when not properly considered. The capacity of anthropogenic activities to modulate the non-stationarity of the hydrological regime, as well as the observed hydrological impacts of these activities, underscores the need for integrated management that considers both natural and human factors, along with their interactions and compensatory effects.